Closing the Gap: Achieving Better Accuracy-Robustness Tradeoffs against Query-Based Attacks

Abstract

Although promising, existing defenses against query-based attacks share a common limitation: they offer increased robustness against attacks at the price of a considerable accuracy drop on clean samples. In this work, we show how to efficiently establish, at test-time, a solid tradeoff between robustness and accuracy when mitigating query-based attacks. Given that these attacks necessarily explore low-confidence regions, our insight is that activating dedicated defenses, such as random noise defense and random image transformations, only for low-confidence inputs is sufficient to prevent them. Our approach is independent of training and supported by theory. We verify the effectiveness of our approach for various existing defenses by conducting extensive experiments on CIFAR-10, CIFAR-100, and ImageNet. Our results confirm that our proposal can indeed enhance these defenses by providing better tradeoffs between robustness and accuracy when compared to state-of-the-art approaches while being completely training-free.

To appear in the Proceedings of the AAAI Conference on Artificial Intelligence (AAAI) 2024

References in corpus (6)

• Explaining and Harnessing Adversarial Examples
• On Calibration of Modern Neural Networks
• Theoretically Principled Trade-off between Robustness and Accuracy
• Countering Adversarial Images using Input Transformations
• Once-for-All Adversarial Training: In-Situ Tradeoff between Robustness and Accuracy for Free
• PopSkipJump: Decision-Based Attack for Probabilistic Classifiers